ÁñÁ«ÊÓƵ

To many people, keen to follow the best advice on how to keep healthy, Kenneth Weiss's book will not be welcome. Surveying a vast expanse of human illness, he argues that much disease is one extreme of normal genetic variation, often unavoidable and in some circumstances an actual advantage. The classic example of this apparent paradox is the haemoglobinopathies, inherited anaemias that protect carriers against the worst ravages of malaria infection. However, the classic example is still the only well-documented example, and to argue the case for the role of evolutionary processes in the development and maintenance of common disorders such as heart disease, Weiss resorts to branches of mathematical biology neither appreciated nor well understood by more empirically minded workers.

Quantitative and population genetics include some of the richest and most successful mathematics in biology, and they are used here to explain how pathogenic genes arise as part of the normal spectrum of genetic variation, without which natural selection, and evolution, could not occur. Most theoretical formulations now accept that a proportion of this variation is selectively neutral, though how large a proportion is still in dispute. Weiss concludes that the neutral model accounts for the bulk of variation (and hence disease), though without warning readers that this is contentious. Workers on genetic variation in the fruit fly Drosophila would offer a different view.

However, the value of the book lies in the mixing of quantitative methods of empirical findings of modern biology. It has now become almost routine to identify the molecular basis of diseases due to abnormalities of a single gene (cystic fibrosis, for example), but in general such disorders are rare. The great challenge now is to deal with common diseases with a genetic contribution: cancer, heart disease, high blood pressure, diabetes, and (although not covered in this book) behavioural disorders. In this endeavour no new technological advances are needed, but instead the application of a dauntingly large number of equations. Most are in Weiss's book.

I have no doubt the next generation of molecular geneticists will be conversant with segregation analysis of quantitative traits and should be thankful for books such as this for introducing the field to them. But they should be warned that they will not find it easy reading, for this is not a book that has grown out of a desire to teach: its purpose is more synthetic than educational.

From arguing that disease is part of the natural pattern of genetic variation, Weiss goes on to ask what makes some genes pathogenic and some beneficial, a question that might be answered from observational data (still very scant) gathered for individual diseases, but one he chooses to answer from the principles of population genetics and the discovery of general principles of biological causation.

This is still a premature venture. Molecular biology continues to reveal the unexpected, such as the very recent discovery that changes in a single gene can destabilise the whole genome and predispose to cancer. The importance of such revelations is that they threaten the assumptions on which much standard population and quantitative genetics rests. For instance, polygenic disease has long been considered the result of many (perhaps hundreds) of genes acting additively, but the first complete molecular analysis of a polygenic disease (diabetes in the mouse) showed that most of the genetic variation was due to alleles at only nine loci and that numerous interactive effects between loci were just as important as additive ones. Discoveries such as this weaken the generalisations of Weiss's arguments.

But in the end, as the work of the quantitative and empirical geneticists converges, a synthesis like that attempted here will prevail. It is too early to say how much that will resemble Weiss's views, but any attempt to introduce more quantitative biology into molecular genetics should be warmly encouraged. It will undoubtedly be fruitful.

Jonathan Flint is a geneticist, Instittute of Molecular medicine, John Radcliffe Hospital, Oxford.